Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
Programming IT Science Technology

First 3D Simulations of Complete Nuclear Detonations 331

jhiv writes: "The National Nuclear Security Administration (NNSA) reports that 'Los Alamos and Lawrence Livermore national laboratories have completed the first full-system three-dimensional simulations of a nuclear weapon's explosion'. The simulations are two of the largest computer simulations ever attempted, each taking weeks to complete on the ASCI White supercomputer. The Los Alamos team used the ASCI Blue Mountain supercomputer to visualize the results. Additional coverage can be found in this story in the Albuquerque Journal."
This discussion has been archived. No new comments can be posted.

First 3D Simulations of Complete Nuclear Detonations

Comments Filter:
  • will this work? (Score:2, Insightful)

    by quinto2000 ( 211211 )
    i remember a great article in Science about 4 years ago that heralded the coming of age of computer simulations as a replacement for nuclear testing. Instead today, the US is trying to back out of the Comprehensive Test Ban Treaty, nuclear weapons are proliferating instead of being disarmed, and if I recall correctly we still test nuclear weapons underground. When will we realize that nuclear weapons are a menace? When will we accept that we need to take the lead in ending their use?

    As an American citizen, I am sometimes disgusted by our government. I really hope that computer simulations can replace the war games, but right now I'm not so certain.

    • Re:will this work? (Score:5, Insightful)

      by Henry V .009 ( 518000 ) on Friday March 08, 2002 @10:21PM (#3133799) Journal
      We do not test nuclear weapons underground, or above ground for that matter. America is confident in its computer simulations. But our confidence in our simulations is not the only factor. Nuclear weapons are never meant to be used. They are meant to deter (threaten). Therefore what is paramount is our enemies' confidence in our simulations.
      • haha, no. they were meant to be used. it was probably sometime during the 1960s that deterence was seen as a viable alternative.

        Kennedy played his cards right and set a precedent.

        What I am personally afraid of is some smaller
        nation obtaining nuclear warfare weapons and using them w/o the understanding of their true results. Not every war could be avoided by ignoring official memos from a drunken leader.
    • Re:will this work? (Score:2, Insightful)

      by IsaacW ( 543020 )
      This is an important result, if only because it simulates the physics of an event in 3-D for a rather complex situation. Many major improvements in technology have come about simply because someone needed a better way to kill someone, and these technologies often find peaceful applications as well. Remember that the early work with nuclear reactions was solely to create a bomb, and from that research we now have safe nuclear reactors that produce very cheap power.

      At the very least, this simulation shows that computers can be used to predict the results of very complex interactions between matter and energy. Surely these same supercomputers can be used to simulate other equally complex phenomena, and these tests break the ice for simulations to come.

      On another note, the United States does not test any physical nuclear devices anymore, underground or otherwise.
    • Re:will this work? (Score:2, Insightful)

      by Combuchan ( 123208 )
      The United States is withdrawing from the Comprehensive Test Ban Treaty because the language in it prevents the signer nations from developing missile-defense anti-nuke shields. The theory was back in 1973 that if one nation had this technology, they could fire their nukes on another nation and be spared from the concept of mutually assured destruction, (MAD) the idea that if you fire one nuke, you essentially end the world.

      The problem with the Test Ban Treaty is that it was written in a different era. Hopefully the events of 11 September indicates to you that there are more than enough people more than willing to kill themselves to inflict as much destruction on the United States as possible.

      MAD doesn't work as an effective deterrent when your enemy is willing to die to kill you.

      I saw a History Channel tagline that referenced nuclear weapons in the most relevant way: "Weapons so powerful, their mere existence implies 'Peace, or else.'"

      This may be straying a bit off-topic here, but what if we hadn't developed the nuke? How many allied and Japanese forces would have died in the invasion of Japan had we not dropped the bomb? 125,000 civillian casualties in Hiroshima and Nagasaki pales in comparison to the amount of possible casualties. The USA estimates 1,000,000 allied casulties in such an invasion, maybe times that by 5 to get the number of Japanese killed.

      Remember. Nukes suck. But they're the better than the alternative. However, straying back on topic ... how do they know this simulation is accurate?

      P.S. If you disagree, don't moderate, reply.
      • The United States is withdrawing from the Comprehensive Test Ban Treaty because the language in it prevents the signer nations from developing missile-defense anti-nuke shields. The theory was back in 1973 that if one nation had this technology, they could fire their nukes on another nation and be spared from the concept of mutually assured destruction, (MAD) the idea that if you fire one nuke, you essentially end the world.

        That's the ABM treaty, a treaty between just the U.S. and Russia. The CTBT is signed by nearly every civilized nation on earth, (including the U.S.) but ratified by very few (also not including the U.S.). We wouldn't exactly have to withdraw from it. (It's not binding since it isn't ratified.)

        Clinton tried to get it ratified in his last couple of years, but had it tabled in the senate because he was going to loose the battle.

      • "MAD doesn't work as an effective deterrent when your enemy is willing to die to kill you."

        Er, the same can be said for missile-defense anti-nuke shields. The danger to the USA isn't from outside the borders: no one's likely to toss an nuke missle at New York. The danger is from within, when some madmen decide to use a suitcase nuke, mail out smallpox, or use cheap'n'easy fertilizer bombs to blow up a federal building.

        The Star Wars shield is just an excuse for funneling endless money to the defense industry; all part of the trend of taking money from the middle-class and give it to the super-rich.
    • Mr. President, We can not afford a supercomputer simulation gap!

      Seriously. We haven't detonated a nuclear warhead since September 23rd, 1992, an underground test called "Divider".

      And I am disgusted by our government from time to time also. But my political views are.. well, too political, even for slashdot.


    • Our government most likely has been doing sims privately for years, ever since the cold war. Now i guess its no longer classified, kinda like the shadow government is no longer a secret.

      As for ending nuclear weapons, we already have better weapons than nuclear, far more dangerous. We have mind control / altering weapons, we have weapons which can destroy all the buildings in an area without harming many people, we have we, weapons which can destroy all the electronics in a city maybe even a small country,we have germ warfare which could kill out millions of our enemies without any of them even noticing we are attacking them, we have the internet which we could use to cause their own people to turn againsnt their governments (if their government lets them on the net)

      Nuclear weapons can do alot of damage but most of the damage is done to the enviornment and to the innocent people, not so much damage is done to the actual military enemy we'd be fighting if they have bunkers and caves and tunnels and bomb shelters. What good is nuke to someone who would survive it? Now germ warfare, and taking down all their electronics with one attack, and bombs which are made to destroy all their underground bases.

      My point is, Nuke alone isnt useful, the only people interested in nuke are terrorists who want to kill innocent people.

      If india were to use nuke on pakistan, lots of innocent people in pakistan would die, the military would launch nuke and many people in india would die, both countries would get absolutely no where.

      Real wars are about missions, targets, and just throwing bombs around is not something thats usually done.

      We threw a bomb at japan but it was because back then no one knew what nuke was.

      Everyone knows about nuke now.

      I worry more about the weapons which we dont know much about, or anything about at all, considering all the technology we have today, the weapons could be way more deadly and a country could be destroyed overnight by something like a virus which spreads through the air.

      I worry about airborn unstoppable virii which kills people in a matter of hours.
      • My point is, Nuke alone isnt useful, the only people interested in nuke are terrorists who want to kill innocent people.

        The nuclear bomb is meant to be a THREAT! not actually used, all this is doing is making sure our stockpile can continue to be a threat.

        If india were to use nuke on pakistan, lots of innocent people in pakistan would die, the military would launch nuke and many people in india would die, both countries would get absolutely no where.

        Precisely. Which is why the nukes will *never* be used in either case. To do so would mean the destruction of either/both country's and massive world commendation. The bombs are simply deterance, or, again a THREAT against an attack and actually probably will make that area more stable in the long run (neither country wants to risk a nuclear war).

        Real wars are about missions, targets, and just throwing bombs around is not something thats usually done.

        Real wars are about survival and winning - without firing a shot. Nuclear weapons ensure the safety of many countrys - not a single country on earth would dare attack the U.S. directly b/c of fear of instant annihilation. Note that Taliban continued to protest its innocence until the very end; and since WWII their hasn't been a single convential attack upon the U.S. A span of 51 years.

        We threw a bomb at japan but it was because back then no one knew what nuke was.

        Ahhh, we knew precisely what the bomb would do - but the japanese didn't. We dropped the bomb to end the war and prevent millions of casulties that would've occured if we had invaded Japan. Even after Hiroshima/Nagasaki the majority of the Japanese war council didn't want to surrender. Only the (supposed) intervention of Emperor Hirohito caused them to surrender.
  • I wanna see the computer simulated results!
    • No problem, here is Wen Ho Lee's phone number...
  • by phr2 ( 545169 ) on Friday March 08, 2002 @10:21PM (#3133798)
    Can we email them the simulation output instead of dropping an actual bomb on them? It would kill a lot less people and be safer for the environment. The opponent could then email us back the results of their own simulation and we'd run a perl script to see which simulation had the bigger yield. It would be like the Star Trek episode where they send people off to the vaporization chambers after checking simulation results, but we could simulate that part too. Wouldn't it be better than the yucky stuff we do now?
  • by bje2 ( 533276 ) on Friday March 08, 2002 @10:22PM (#3133803)
    i thought i already saw someone do a computer sim of a global thermonuclear war....

    oh wait...that was just matthew broderick in was games....
  • Intriguing Photos (Score:5, Informative)

    by EricKrout.com ( 559698 ) on Friday March 08, 2002 @10:26PM (#3133815) Homepage
    There are some really intriguing photographs of nuclear detonations and related events on the web.

    I happened to stumble across one particular site with great photos: http://www.dtra.mil/about/photo_library/ab_combat. html [dtra.mil]

    monolinux.com {Linux news only. No ads.} [monolinux.com]
    • Hmm. This reminds me of something i pondered, once, to no avail.. maybe someone on slashdot will know..

      What is the phenomena, demonstrated in this photo: here [dtra.mil] that causes the streamers in the background? I saw them in almost all the nuclear detenations in "Trinity and Beyond".

      • scale rockets. they fired off little smoke rockets a (very short) bit before the blast to give a physical measure in teh background as to distance of expansion. I've seen a picture where I guess the wind was even less than usual that day where you could still see the diamond pattern almost perfectly behind the (then very small) fireball; diamond becuase in that instance they fired them off in angles. Naturally this isn't persistant more than a fraction of a second, but that fraction of a second probably contains all you're interested in as a bomb designer (becuase after that second, the effects are generally pretty well known ;-) ).
        • You can always find fun things on the fas.org site, like this image [fas.org] rather clearly depicting the scale rocket trails. Also on the same page [fas.org] is a fascinating image (and explanation of) the "surface mottling" and "spike" phenomenon seen in some really-small-time-after-detonation explosion photos [fas.org]. (it's at the bottom of the page)
      • Re:Intriguing Photos (Score:2, Informative)

        by NetRanger ( 5584 )
        The streamers were used for several purposes:

        1) To measure the overpressure effects of the shock wave (a very useful tool to determine optimum blast height)

        2) To gauge the size of the fireball (note the relatively even distribution of the streamers)

        3) To give the capability to exactly duplicate a past test, using film stock as a gauge. You could then measure the exact detonation point of a shot from the streamers from multiple camera angles and simple math.
  • by Ars-Fartsica ( 166957 ) on Friday March 08, 2002 @10:26PM (#3133817)
    How about an accurate simulation of the proliferation of Soviet-era weapons grade fissionable materials?

    This may come across as offtopic, but I strongly believe that is where every red cent of our nuclear program money should be spent.

    You can either find this material now or it can find you later.

  • Vaguely amusing (Score:3, Interesting)

    by Karma Sink ( 229208 ) <oakianus@fuckmicrosoft.com> on Friday March 08, 2002 @10:27PM (#3133822) Homepage
    Meanwhile, the quote at the bottom of the page is "Walk softly and carry a big stick."
  • I'm sure there will be comments about how this is not the best use for the most powerfull computer in the world. Searching for aliens, folding protiens, weather modeling, cataloging the genome, etc.. But this beats the hell out of live testing. We have come a long way from the days of above ground testing next to Los Vegas or bombing the hell out of Bikini Atoll.

    Can't help but think of how many units I could rack up with just a day on this baby though :)
    • I'm sure there will be comments about how this is not the best use for the most powerfull computer in the world.

      I think I heard that ASCI Blue Mountain is also used for designing greeting cards.


      It could be that I'm slightly mixed up. :-)

  • by Graelin ( 309958 ) on Friday March 08, 2002 @10:30PM (#3133827)
    "We've got to move into the future without testing," said Bob Weaver, who headed the Los Alamos effort.

    ...sounds like my boss...
  • They said they compared the simulation with an actual underground test. How did it measure up? The article didn't say.

    Or is that secret?
  • 750 years? (Score:2, Interesting)

    by Erris ( 531066 )
    The AJ article had an estimate of 750 years of run time for a good home computer to do the same thing. So, how long would it take for a few thousand home computers, good, bad and ugly? Do you know what that cute little screen saver is really doing? Bwa-ha-ha-haaa!
    • Assuming the job is easily distributable, without things like large bandwidth requirements between nodes, then it's just arithmatic. 750 home computer years, 750 home computers take around one year, plus a year or two for efficiency loss. Now you know why the college computer labs freak out when you install things like dnet or seti on them.....

      They just don't want to have to comply with the "military grade computer" requirements! :)
    • Currently, the UD molecule screener [ud.com] uses as much as 375 CPU-years per day.

      On that network you could run this nuke simulation twice a week. Without building a billion-dollar computer.

      Of course, you're probably not going to get the same amount of participation with ATOM BOMB research as you are with CURING CANCER research (Do you know what that cute little screen saver is really doing? Bwa-ha-ha-haaa! -Erris [slashdot.org]). But overt ANTHRAX VACCINE research got a decent fraction of the way there, and is closer to the former than the latter, in practical terms.

      --Blair
  • by T5 ( 308759 ) on Friday March 08, 2002 @10:49PM (#3133894)
    By law, the directors of the nuclear weapons labs (Sandia, Los Alamos, and Lawrence Livermore, IIRC) are required to certify annually the readiness of the nuclear stockpile. This has been a problem due to the lack of production of tritium in the US, with the exception of a small amount from Savannah River in South Carolina and just recently at the Tennessee Valley Authority's Watts Bar Nuclear Plant in Tennessee. Without sufficient quantities of tritium, the aging thermonuclear arsenal's decay of tritium puts the existing weapons at risk of not functioning within their design parameters. The only thing more frightening than a nuke that works is one that you can't rely on to work when needed. Thus, the directors have threatened to not certify the arsenal.

    With this new computing power, the directors can now verify the status of even degraded weapons whose functionality was up till now a mystery and make better decisions about how to use the still small amounts of tritium being produced.
  • by Anonymous Coward on Friday March 08, 2002 @10:52PM (#3133903)
    First lets figure out how many times faster this computer is.

    One of them they gave statistics that it did the work of 750 years worth of computer time in 39 days.
    First lets figure out how many days there are in 750 years
    750*365.25(accounting for leap year) ~= 273937

    Thats 273,937 days in 750 years (give or take a couple of weeks)

    Now 273,937 / 39 gives us are actual ratio which is a factor of 7024.
    This means that the los Alsomething is 7024 times as fast.

    Now a typical computer now a days can run quake 3 at around an average fps of 60.
    7024 * 60 gives us the fps of the super computer.

    Which is a grand total of 421,422 FPS!!!!

    My only question... When can I buy one?

    • You can't.

      As far as I know, the ASCI White system is pretty much "classified" in a way. I know it uses Power3 processors... over 8000 of them, in fact. And it has a mesh-type interconnect... very expensive stuff here. You can see how it compares here [top500.org]

      Note that the second most powerful supercomputer peaks at HALF the GigaFlops... there's some serious power there. Though that list is technically innacurate... the #2 machine only has 3000 processors, and 8 on standby in case of hardware failure. That one is an "open" cluster, however. I don't think it's opened quite yet, and I can't even remember the name... something French... LeMue or somesuch.

      I did take a look at it once, though. Sweet machine(s) :-) And you can't help but be impressed when you learn that it took 4.5 TONS of cabling to wire that puppy!

      • I did take a look at it once, though. Sweet machine(s) :-) And you can't help but be impressed when you learn that it took 4.5 TONS of cabling to wire that puppy!


        That's less impressive if you've every worked with cabling those high end systems. 4.5 tons of Craylink would only be like 100 feet (at least when you have to carry it around all day. :)

        It's not quite that bad, but when you have 500 conductor cables the weight does add up quick.

  • I'm sure our government has done it 20-30 years ago but its been classified.

    If anyone honestly believes its the first EVER and that our military didnt do this kinda stuff during the cold war they are crazy.

    trillions of dollars have been spent over the last serveral years, 10s of trillions over the last 20 years and this is the first sim, in 2002? hahaha a joke right?
    • NO this is a first. Maybe simulations into the effects of a nuke were done but never and actual simulation of one going off.

      Just think about how much physics is needed for this. Hudge temperature gradients, wavefronts the list is endless. This all has to be simulated.

      We can't even simulate weather properly yet becasue of lack of computing power.

      • So you have top secret military clearance? how do you know? Oh wait because they said so? you believe everything they said?

        You cant believe what the government says, classified means they wont tell you about it or will lie to you until they are read to announce it.

        Its announced now, but it could have been simulated in a classified fashion in the past.
        • You cant believe what the government says, classified means they wont tell you about it or will lie to you until they are read to announce it.

          The question is how did they have the computer power? Even for the military, these computer prices are a bit expensive, and with Moore's law, we haven't had the computing power available at any cost until recently.
          • Not to increase conspiracy paranoia, but it's entirely possible that the government has their own completely seperate design, fab, etc on a whole line of top secret processors that is all top secret. And also, who's to say that the government hasn't approached Intel, and convinced them to lag their releases one generation behind what they make available to the government, all top secret.

            That's the thing about secrecy, you never know!

            It's not likely, but it is possible. BTW- Moore's law is no more a law than Murphey's law. It's just a prediction that is uncannily true so far.
            If we stopped developing all new processors, I guess that would disprove Moore's law? Suppose AMD hadn't made an end run on Intel. do you think processors would be as fast as they are today? My point is, the rate of processor speed (or transistor density for the purist), isn't dictated by any law, it's dictated by how much effort we put into it, which is dictated in part by market forces when you are tlaking about corporations.
            • Not to increase conspiracy paranoia, but it's entirely possible that the government has their own completely seperate design, fab, etc on a whole line of top secret processors that is all top secret.

              A multi-billion dollar project - so we can see how a nuke explodes? It makes much more sense to use that money to keep bases open or build more planes and warships, considering how tight the military budget has been recently.

              That's the thing about secrecy, you never know!

              Ergo, cognito sum. We truely know almost nothing. Rational deduction from sensory input can lead us far, though.
  • Hmmm... (Score:2, Funny)

    by andyring ( 100627 )
    Those computers are extremely cool (I want one!), and hey, if they do the job without having to blow things up, it works for me. But, how long until some Microsoft salesweasel comes along and tries to convince them to run NT on it? Brings new meaning to your computer bombing....
  • Has squaresoft announced a port for Final Fantasy XI on the ASCI White computer yet? Although I guess a simulation of nuclear destruction is about as final a fantasy as you can get
  • First, ASCI White specs: Name: ASCI White Built by: IBM over a period of 5 years for the Department of Energy Price: for $110 million. Power: 1,000 times more powerful than IBM's Deep Blue; capable of roughly 12.3 trillion calculations per second CPU: Made of off-the shelf IBM Power3 processors (well, 8,192 of them altogether) RAM: 16 terabytes Disk space: 160 terabytes Power requirements: 3 megawatts of electricity (would light up 3,000 homes) Now let's say 1 Power3 = 1 domestic processor of today... By Moore's law, in 18 months our computational power will double, so: 2^x = 8192 x = 13 13 * 18 months = 19,5 years conclusion: video games will be very cool in 20 years
    • Deep Blue is a stupid comparison. The ASCI series is built with general purpose processors. ASCI Red was made with Pentium 200s if I recall correctly. Something like 6 or 8 thousand of them.

      Deep Blue used ASICs to boost operations that were specific to chess. Anyway, I guess their PR department needed an easy comparison.
  • addendum... (Score:2, Funny)

    by Loki_1929 ( 550940 )
    Several hours later it was discovered that the software used by the Los Alamos and Lawrence Livermore national laboratories had a slight flaw, and the corrected simulations show that the nuclear explosions were in fact beige. [space.com]

    In a related story, the updated software was found to contain massive amounts of spyware. [newsbytes.com]

  • F*ck this, I won't read all the crap. In 1963, the Soviet Union and the USA signed the Limited Test Ban Treaty. This banned atmospheric as wel as some other testing. Underground testing is not banned. In 1975, the Threshold Test Ban Treaty is passed that limits underground testing to 150KT. Now comes the comprehensive test ban treaty.

    Okay, if you use google, then you could likely find all of this. You are karma whores (maybe).

    In response to the the LTBT, the USA (and others) implemented the WWSSN to help monitor nuclear testing (go to google and type in WWSSN). I looked at a lot of WWSSN records from TLA WWSSN sites).

    After the TTBN treaty is passed, the USA accuses the USSR of cheating; the axle (note, one hub) of evil is detonating bigger bombs than 150 KT. It is later recognized by scientists that bombs detonated in Nevada will produce a bigger recorded seismic signal that those in Novya Zemlya. Nevada has more highly attenuating rock in the area, Novya Zemlya is more solid bedrock. About ten years later it is finally believed that the Ruskies didn't cheat.

    It has been speculated and investigated that a rogue nation could detonate a big mother bomb in an evacuated cavern. This "hollow" bomb chamber would severely attenuate the outgoing seismic signal. Hmmm, dig a hole that could hold the Statue of Liberty...

    Got to go to sleep before I ramble all night.
  • Kinda scary, isn't it, that 750 years of current desktop runtime is one year of desktop runtime in 15 years, according to Moore's law?

    Don't let that program get out... especially on a "misplaced" hard disk...
    • What use would it be to anyone? Even if a rival country got it, it probably would only be useful for simulating OUR bombs, not theirs.
  • As I was reading the details to this story on ABQJournal, I saw this odd link on the bottom of the page.

    (PRC# 3.4676.620668) [icopyright.com]

    Wondering what a PRC article was, I followed it, only to discover various ways I couly pay to use this article.

    One of the options is "HTML Link : Publisher's permission to link to an article." and they charge $50 for this privledge. Apparently this whole scheme is managed by iCopyright.com

    I thought there was some legal precedence in the past allowing deep linking. What happened?

    • This is great.

      It's probably just trolling (not the way slashdot people use trolling).

      Icopyrgight.com probably just fishes for suckers to pay the fee. If someone links to the site who is a little guy, they will just pay the $50 rather than try to fight it in court. Inethical business model, hope to see them sued into oblivion. Bastards.
  • Look like the BFAs [twangdemonium.com] are here. This is the seond time I've seen one. Not quite time to subscribe, though

  • Hey, everyone should email Los Alamos and see if they'll GPL the code. Wouldnt' that be great!? ;)
  • Since /. ran the TRON review yesterday: Livermore Labs is where the "big door" leading to the ENCOM labs is located.

"Being against torture ought to be sort of a multipartisan thing." -- Karl Lehenbauer, as amended by Jeff Daiell, a Libertarian

Working...